The proposed study aims to uncover the predictive relationship between physiology and violent behavior in adolescents who are, and are not at social risk. Autonomic as well as cortical measures will be examined both at baseline, and in response to stressor and challenge tasks. These measures will be compared in groups of adolescents who have both social and biological risk factors, biological only, social only, or controls. It is hypothesized that adolescents with no social risk factors but very low physiological arousal will be more likely to become violent than subjects with normal levels of arousal. Additionally, it is suggested that adolescents with social risk factors but with high levels of arousal will not be as likely to become violent as at-risk subjects with normal levels of arousal. This study will allow for the examination of adolescents who are at high social risk but are protected from violent behavior through physiology. The examination of the interaction between biological, social, and psychological factors, such as impulsivity, will thus be possible. These measures will have implications for the role biology plays in putting adolescents at risk for violent behavior as well as its ability to protect them from developing this behavior.